Rationale for Transfer Modelling

Scaling up a pharmaceutical process or transferring it between equipment platforms is inherently challenging. Even when formulation composition remains unchanged, differences in geometry, dynamics, and operating ranges between small- and large-scale devices can lead to significant shifts in critical quality attributes (CQAs).

Collecting extensive experimental data on large-scale equipment is often costly, time-consuming, and material-intensive. In contrast, small-scale devices, such as compaction simulators or laboratory-scale presses, allow efficient and systematic data generation. Leveraging these advantages requires modelling strategies that can transfer knowledge from small-scale systems to larger-scale ones.

Role of Strong Core Models

Transfer modelling at Elegent builds on the existence of strong, well-validated core models. These core models describe the relationship between:

• Raw material properties
• Formulation composition
• Process settings
• Resulting CQAs

and are typically developed using data from small-scale, highly instrumented devices.

Because these models capture the fundamental material–process–response relationships, they provide a robust starting point for model-building at larger scale. Rather than constructing large-scale models from scratch, the core models serve as a baseline representation of formulation behavior.

Efficient Use of Large-Scale Data

To adapt a core model to a larger-scale device, a limited amount of carefully selected large-scale data is collected. This data is used to retrain or fine-tune the existing model, allowing it to account for:

• Differences in machine geometry
• Changes in dwell time or compaction dynamics
• Scale-dependent effects on flow, compaction, or ejection

Because much of the underlying structure of the model is already learned at small scale, only a relatively small dataset is required to achieve reliable predictions at large scale.

This approach contrasts with traditional scale-up strategies, which often rely on extensive experimental campaigns to rediscover relationships that are already well understood at smaller scale.

Benefits for Scale-Up and Process Transfer

By combining strong core models with targeted large-scale retraining, transfer modelling supports:

• Faster and more reliable scale-up
• Reduced experimental burden on production-scale equipment
• Improved prediction of CQAs during process transfer
• Earlier identification of scale-related risks

In practice, this enables formulation and process development to progress with greater confidence when moving from laboratory or pilot scale to commercial manufacturing.

Summary

Transfer modelling provides a structured pathway to extend predictive CQA models from small-scale to large-scale equipment. By anchoring large-scale models in robust small-scale core models and complementing them with limited, efficiently chosen large-scale data, Elegent’s approach supports scalable, data-efficient, and predictive process development.